News & Analysis
Comment
resistion
China's interest in PCRAM is not strong. Its interest in RRAM is much stronger.
hm
Since Microchip's Sanghi is interested in this technology and new products, ...
China develops phase-change RAM
Peter Clarke
4/20/2011 7:55 AM EDT
LONDON – The Shanghai Institute of Microsystem and Information Technology (SIMIT) at the Chinese Academy of Sciences – working with foundry chipmaker SMIC and Microchip Technology – has announced it has developed phase-change random access memory (PCRAM) that is based on Chinese intellectual property, according to a Xinhua report.
The 8-Mbit memory, scheduled for mass production later this year, is intended for use as a replacement for NOR flash in applications such as mobile storage and RFID, according to Chinese language reports in translation. It is relatively modest in capacity, compared with available flash memory but for China the development is intended to break the foreign control of memory chip production and has been accompanied by the filing of about 200 patents, the reports said.
The PCRAM was manufactured jointly by Shanghai Institute of Microsystem and Information Technology of the Chinese Academy of Sciences, Semiconductor Manufacturing International Corp. and Microchip Technology Inc., according to Xinhua, the official press agency of the government of the People's Republic of China. No details were provided of the geometry of the manufacturing process or plans to scale the manufacturing or produce devices with increased capacity.
While it is expected that SMIC could act as a foundry manufacturer of the relatively modestly sized memory it is not yet clear what the role of Microchip (Chandler, Arizona) is in the development of the memory.
Phase-change memory – like NOR and NAND flash – is nonvolatile, providing power-saving opportunties. Its operation is conventionally based on the change of resistance as a chalcogenide glass changes state from amorphous to crystalline under the heating effect of a current passing through a thin film. Although theoretically superior to DRAM, SRAM and flash memory on a number of counts it has yet to reach mainstream commercial adoption.
Micron and Samsung are in the lead. Samsung has developed a 512-Mbit PCRAM test chip on a 60- to 65-nm manufacturing process which has achieved some deployment in Samsung mobile phone. Micron, through its Numonyx acquisition has a 128-Mbit device on 90-nm process. Numonyx was also working on a 1-Gbit phase-change memory on a 45-nm process.
The Shanhai Institute's PCRAM test chip was made on 200-mm diameter wafers but details of the manufacturing process technology or whether it is based on chalcogenide material were not discussed in Chinese language reports.
A voice demonstration has confirmed that the chip can read, write and erase, according to Xinhua. The patents cover technologies ranging from the distribution of materials, structures and chip design to means of testing. China plans to be able to supply 60 percent of its own memory IC requirements within 10 years.
Related links and articles:
Xinhuanet article
Samsung CEO: Headwinds hinder PRAM
Update: PCM found in handset
CTO confirms IBM's PCM expectations
A series of articles on PCM authored by Ron Neale and published in EE Times Memory Design Line:
PCM Progress: Temperatures rise and constituents on the move
PCM The Myth of Scalability Part 3- Is WAL-PCM the salvation?
PCM Scalability:The Myth (Part 2)
PCM scalability--Myth or realistic device projection (Part 1)
|
|
|
|
|
Navigate to related information
unknown multiplier
4/20/2011 10:11 AM EDT
If it's not from Ovonyx, that would be interesting.
Sign in to Reply
Volatile Memory
4/20/2011 11:31 AM EDT
This is the funniest article I have read in a while. Mr. Clarke has done it again!
The fact is, Samsung has achieved no deployment in any mobile phone. Samsung just floated a few fake, non-commercial units of the low-end GT-E2550 model, before it realized that the chips have power-consumption issues, immediately revised the bill of materials for the model, and replaced the chips with plain NOR MCPs last summer.
Micron, of course, has no sales of its undeperforming and overpriced 128-Mbit chip and is not working on any 1-Gbit chip.
Shanhai Institute's PCRAM test chip may indeed read, write and erase - up to 1000 times, at the speed of a snail.
Ovonyx, by the way, will file for Chapter 11 by the end of 2012.
Sign in to Reply
Volatile Memory
4/20/2011 11:37 AM EDT
One more thing - surely the Chinese are gunning for rad-hard chips, after their latest shipment was confiscated. Unfortunately, the Chinese failed to read the JPL report which shows that these PCRAM chips throw off errors like crazy. And BAE has persistently failed its US military qualifications for these rad-hard PCRAM chips since 2006. Now it is offering them to the Russians, hoping to ruin their space program.
Sign in to Reply
resistion
4/20/2011 11:36 PM EDT
I think Chinese Academy of Sciences used chalcogenides like Ovonyx, based on the SIMIT papers I found. If these materials melt more easily than other semiconductor industry materials, their structure cannot be that stable..
Sign in to Reply
peter.clarke
4/21/2011 5:00 AM EDT
@ubm112211
Perhaps you are referring to Shandong Sino-chip Semiconductors Ltd.? But as to the power which buried Qimonda; you will need to be more specific, without being libelous.
Sign in to Reply
peter.clarke
4/21/2011 5:11 AM EDT
@resistion
I also suspect that SIMIT is using chalcogenide mixes and the assertion that the development is entirely proprietary is a legal position. If the development is ever commercially successful one might expect some patent horse trading or legal action, eventually.
Sign in to Reply
JackT
4/21/2011 6:43 AM EDT
China Gov: Funding
SIMIT: Material/Phyics
SMIC: Fab manufactory/Process Intergration
Microchip SST: TD/Process/Device/Design/Product/Test/Marketing
SST was acquired by microchip last year
See the link: http://english.sim.cas.cn/ns/icn/200910/t20091019_45466.html
Volume production? Is it NOR compatible product? Interesting and waiting for their risk production anoucement.How can they do something better than Micron.
Sign in to Reply
vincentxia
4/21/2011 11:57 AM EDT
I worked in SIMIT before, known every single detail about it.
the purpose of this research is not for mass production, but just getting funding from the gov.
Sign in to Reply
peter.clarke
4/21/2011 12:15 PM EDT
@JackT
Thanks or the insight and the URL
@vincentxia
Some would argue it was ever thus within research groups; in the commercial world as well as in academia; in the west as well as in the east.
Sign in to Reply
JackT
4/21/2011 9:28 PM EDT
Microchip/SST split gate flash cell may don't have road map below 65nm.They may think something diffrent. That is the reason.
They may has some sigma data to show the performance, but it really high risk to put in into MCU even their low end PIC market.
Sign in to Reply
resistion
4/21/2011 10:07 PM EDT
Some companies have NOR flash roadmap below 65 nm. At least to 4X nm. Beyond that is uncertainty for sure.
Sign in to Reply
hm
5/8/2011 8:32 AM EDT
Since Microchip's Sanghi is interested in this technology and new products, there will be specific market requirement for PCRAM.
Sign in to Reply
resistion
5/8/2011 11:18 AM EDT
China's interest in PCRAM is not strong. Its interest in RRAM is much stronger.
Sign in to Reply